hurricane ivan damage survey

8/9/2019 Hurricane Ivan Damage Survey

1/12

P 7.4 HURRICANE IVAN DAMAGE SURVEY

Timothy P. Marshall*

Haag Engineering Co.

Dallas, Texas

1. INTRODUCTION

The author conducted aerial and ground damage

surveys along the Florida and Alabama coasts after

Hurricane Ivan. The purpose of these surveys was to:

1) determine the height of the storm surge, 2) acquire

wind velocity data, 3) determine the timing of each, and

4) assess the performance of buildings exposed to wind

and water effects. Particular emphasis was placed on

delineating wind and water damage. A similar study

has just been published by FEMA (2005).

The author rode out Hurricane Ivan near Pensacola,

FL then conducted hundreds of site specific inspections

the year following the hurricane. Most buildingsexamined were wood-framed structures. Remaining

buildings consisted of concrete masonry as well as

multi-story, steel-reinforced, concrete structures.

Various building failure modes were observed.

Typically, wind exploited poorly anchored or attached

roofs and vinyl siding whereas wave action

undermined, collapsed and destroyed buildings near the

coast. Wind damage generally began at roof levels

whereas wave damage attacked the bases of buildings.

Both lateral and uplift forces were applied to the

buildings from wind and water and examples of such

failures will be shown in this paper. Delineating the

damage between wind and water involved knowledgeof building construction, as well as understanding the

direction and magnitudes of the wind and water forces

during the hurricane. A primer on the subject had

been published by FEMA (1989).

2. WEATHER BACKGROUND

Hurricane Ivan struck the Alabama coast and

western Florida panhandle during the late evening on

September 15, 2004 and early morning on September

16, 2004 (Fig. 1). According to Stewart (2005), Ivan

weakened as it approached the coast due to a number of

environmental factors and dropped to category 3

strength on the Saffir-Simpson Scale. Refer to Table 1.

The eye of the hurricane made landfall near Gulf

Shores, AL around 0700 UTC (2 a.m.) local time and

the eyewall tracked northward up the Perdido River

along the Alabama/Florida state line.

______

*Corresponding author address: Timothy P. Marshall,

4041 Bordeaux Circle, Flower Mound, TX 75022-7050.

email: [email protected]

Figure 1. Enhanced color infrared satellite image of

Hurricane Ivan at landfall near Gulf Shores, AL around

0700 UTC (2a.m.) on the morning of 16 September

2004. Arrow indicates location of author. Image

courtesy of NOAA/NWS.

Analysis of radar data revealed that Hurricane Ivan

had a closed eyewall until it was about 100 km (62

miles) from the Alabama coast. According to Stewart

(2005), a combination of dry air from Louisiana,

upwelling of cooler water near the coast, and increasing

wind shear from an approaching upper trough haddetrimental effects on storm strength. As a result, the

southern half of the eyewall eroded away just prior to

making landfall and surface wind speeds decreased

(Fig. 2). The north eyewall crossed the Alabama coast

around 0600 UTC (1 a.m.) near Gulf Shores, AL with

the east eyewall passing over Perdido Key, FL. The

eye crossed the coast about an hour later. By 0800

UTC (2 a.m.), the eyewall extended from Mobile, AL

to just west of Pensacola, FL.

TABLE 1

SAFFIR-SIMPSON SCALE

NO. WIND* (mph) SURGE (ft)

1 74-95 4-5

2 96-110 6-8

3 111-130 9-12

4 131-155 13-18

5 >155 >18

*sustained wind (1 minute average)


2/12

Figure 2. Radar image of Hurricane Ivan at 0641 UTC

(141 a.m. local time) on 16 September 2004 as it

approached the Alabama and Florida coast. Noteerosion of south and west eyewall. Image courtesy of

NOAA/NWS.

2a. WIND SPEEDS AND DIRECTION

The strongest winds from Hurricane Ivan occurred

in the north and east eyewall which passed just west of

Pensacola, FL. At 0638 UTC (138 am), the Pensacola

Naval Air Station reported sustained winds of 39 ms-1

(87 mph) with a gust to 48 ms-1

(107 mph). At 0644

UTC (1:44 am), the Pensacola Airport reported a wind

gust to 47 ms-1 (106 mph) from the east-southeast at 10

meters (33 feet) above the ground in open, unobstructedterrain. A Doppler on Wheels (DOW) truck in Gulf

Shores, AL also measured a wind gust of 51 ms-1 (115

mph). Wind speeds were lower east, west, and inland

of the east eyewall. Initially, the wind direction was

from the east along the coast. Then, as the eye made

landfall, winds east of the eye shifted gradually to the

south.

The Florida Coastal Monitoring Program (FCMP,

2004) also had wind measuring stations in several

places near the coast. Wind velocities weakened at the

Fairhope, AL site indicating that the eye passed over

this location. Also, there was a sudden change in wind

direction from east-southeast to west-southwest as the

eye passed. In contrast, the Pensacola site never

experienced the eye (Fig. 3). Winds in Pensacola

shifted more slowly from east to southeast to southwest

as the eye passed to the west. A peak 3-second gust of

just over 49 ms-1

(111 mph) was recorded at 0643 UTC

(143 am).

Figure 3. Wind speed (mph) and direction from

Pensacola during Hurricane Ivan. Figure courtesy of

the Florida Coastal Monitoring Program.

2b. STORM SURGE

The storm surge precedes and accompanies a

hurricane. This occurs as the hurricane pushes seawater

ahead of it. At the same time, the hurricane moves

towards shore. The coast acts as a barrier to the rising

sea levels resulting in a squeeze play where water isliterally pushed onto land. Waves are superimposed on

top of the storm surge. As indicated by Simpson and

Riehl (1981), the peak storm surge occurs east of the

eye and is typically coincident with the peak winds.

The author measured the height of the storm surge

using a surveyors level and rod. Still water lines in

buildings provided the best estimate of the storm surge

level. The line was formed by dirt and debris in the

water that was deposited on wall surfaces. Generally,

the still water level was found in a room that was not

breached by wave action. Occasionally, a line of grime

was found deposited on glass items or rust on metal

items due to contact with salt water. Sometimes scrape

marks were noted in wall surfaces from impacts by

floating furniture (Fig. 4).


3/12

Figure 4. Indications of the height of water: a) dirt line

in bathroom, b) rust line on metal fuse box, c) scrape

marks on paneling, and d) scrape marks on trees.

In the absence of a building, scrape marks on trees

provided a measure of water depth. The scrape marks

were formed by repeated impacts of floating debris

which abraded and removed the bark. The heights of

the scrape marks usually were close to heights of the

still water lines provided the debris remained in contact

with the trees. Flotsam and debris lines also provided

an estimate of the height of the storm surge. The author

obtained dozens of still water line measurements from

various locations extending as far west as Ft. Morgan,

AL and as far east as Fort Walton Beach, FL. Selected

measurements appear for the Florida coast appear in

Figure 5.

Figure 5. Selected measurements of still water heights

(ft.) above sea level for the Florida coast after

Hurricane Ivan.

The highest storm surge measured was 4.4 m (14 ft.)

at Perdido Key, FL. Storm surges in excess of 3.8 m

(12 ft.) were measured in Grande Lagoon, Escambia

Bay, Tiger Point, and Blackwater Bay areas. Low-

lying areas along the coast (i.e. Grande Lagoon and

Tiger Point) were subjected to the full force of themoving water including wave action and sustained the

greatest concentration of damage. Barrier islands,

which normally protect the mainland coast, were

submerged during the storm. Bays that faced south

channeled the water such that the north end of

Escambia and Blackwater Bays had storm surges nearly

equal to the highest levels on the coast. The highest

surge (including wave runup) found was 5.6 m (18 ft.)

in the Seaglades subdivision southwest of Pensacola,

FL.

2c. TIMING OF WIND AND WATER

Wind and water data were compared for three sites:

Dauphin Island, Pensacola, and Panama City. In each

instance, the storm surge exceeded normal high tide

during the early morning of September 15th

, about 24

hours before the eye made landfall. The storm surge

rose steadily throughout the day then increased rapidly

as the stronger winds moved ashore. The peak storm

surge coincided with the strongest winds (Fig. 6).

Figure 6. Wind speed (knots), wind direction and

water height (ft) from Dauphin Island, AL during

Hurricane Ivan. Yellow line indicated sustained winds,

and the vertical red line indicated peak gusts. Wind

direction is the blue line and water height is the red

line. Courtesy of the National Ocean Survey.


4/12

3. WIND SPEED-DAMAGE CORRELATION

Mehta et al. (1983) correlated wind speeds with

building damage after Hurricane Frederic. Varying

degrees of building damage were assigned failure windspeed values depending on the degree of engineering

attention to the building.

McDonald (2005) further advanced the concept of

wind speed-damage correlation by assigning failure

wind speed ranges based on the degree of damage

(DOD) to 28 types of buildings and objects. For wood-

framed residences, McDonald indicated that the

removal of roof coverings generally occurs with a

three-second wind gust of about 36 ms-1 (80 mph).

The removal of the roof deck occurs with a three-

second wind gust of about 44 ms-1

(98 mph), and the

removal of the roof structure occurs with a three-second

wind gust of 54.5 ms-1 (122 mph). Variations up to 20

percent can occur depending on the type of building

construction and the extent of anchorage. Also,

building items not damaged would give an upper bound

failure wind speeds. In this study, wind speed-damage

correlations were determined for selected locations and

the results are shown in Figure 7 along with actual wind

speed measurements.

Figure 7. Actual and estimated 3-second peak wind

gusts (in mph) at 10m (33 ft.) above the ground in open

terrain for Alabama and Florida from Hurricane Ivan.

Actual values (circled) are from the Florida Coastal

Monitoring Program. Estimated values are based on

wind speed-damage correlations.

Actual and estimated three-second peak wind gusts

from Hurricane Ivan were then compared to the design

three-second gusts as stated in the ASCE 7-95 (1996)

standard (Fig. 8). This standard indicates that structures

built along the coast should be designed for 130 mph

three-second gust. It was found that the Hurricane

Ivans winds were lower than those stated in the ASCE

7-95 standard.

Figure 8. Comparison between the basic design wind

speeds in ASCE 7-95 (black lines) with those from

Hurricane Ivan (red lines) for Alabama and Florida.

Wind speeds are in mph with ms-1 in parentheses.

4. DAMAGE BY COMMUNITY

The following is a summary of damage observations by

community.

4.1 FORT MORGAN, AL

Fort Morgan was located at the west end of

Highway 80 on the east side of Mobile Bay. The area

experienced the northern eyewall with strongest winds

from the east. Wind damage to buildings was minimal

and consisted of damage to asphalt shingles and vinyl

siding. In general, well-built homes sustained little or

no wind damage. Maximum wind gusts at 10m (33 ft.)

were estimated to be around 45 ms-1 (100 mph). The

storm surge was around 3 m (9.6 ft.) above normalocean water level. Up to 1 m (3.2 ft.) of sand was

transported into the first floors of coastal homes.

4.2 FORT WALTON BEACH, FL

Fort Walton Beach was located well east of the

eyewall in Hurricane Ivan and experienced the

strongest winds from the east. Wind damage to

buildings was minimal and consisted of damage to

asphalt shingles and vinyl siding. In general, well-built

homes sustained little to no wind damage. Maximum

wind gusts at 10 m (33 ft.) in open terrain were

estimated to be around 42.5 ms-1 (95 mph). Lowerwind speeds occurred in forested areas. Storm surge

and wave action inundated the first floors of oceanfront

homes and destroyed swimming pools and patio decks.

4.3 GULF BREEZE, FL

Gulf Breeze was located east of the eyewall and


5/12

experienced the strongest winds from the east-

southeast. Some pine trees were uprooted or snapped to

the northwest. Wind damage to buildings was minimal

and consisted of damage to asphalt shingles and vinyl

siding. In general, well-built homes sustained little to

no wind damage. Maximum wind gusts at 10 m (33 ft.)

in open terrain were estimated at around 49 ms

-1

(110mph). Wind speeds were significantly lower in

forested areas.

Storm surge and wave action were severe and

reached between 3.1 and 3.4 m (10 to 11 ft.) above the

normal ocean water level on the Santa Rosa Sound side

and between 2.6 and 2.8 m (8 and 9 ft.) on the bay side.

Waves were superimposed on the storm surge and

severely damaged or completely destroyed those

structures on grade at low elevations. One of the worst

hit areas was Tiger Point subdivision where houses

along Santa Rosa Sound were washed away. Trees and

other vegetation were killed by saltwater inundation.

4.4 GULF SHORES, AL

The eye of Hurricane Ivan passed over Gulf Shores

with the strongest winds occurring from the east in the

northern eyewall. Wind damage to buildings was

minimal and consisted of damage to roof coverings and

vinyl siding. However, occasional damage was

observed to east gable ends and some roof decking was

removed. Roof damage occurred when wind was able

to get underneath an overhang or enter the building

through a broken window or door. In general, metal

roofs performed better than three-tab shingle roofs.

Well-built homes sustained little to no wind damage.

Maximum wind gusts at 10 m (33 ft.) estimated around51 ms

-1(115 mph) in open terrain.

Storm surge ranged between 3.1 and 4 m (10 to 13

ft.) above the normal ocean water level from the west

end of the island to the east end. Buildings elevated

above the storm surge escaped significant damage.

However, buildings were severely damaged or

destroyed when waves reached the second story.

Floating debris battered and mangled pilings. Up to 2.5

m (8 ft.) of sand was removed from the beach and

transported inland covering the main highway, Rt. 182.

Pilings that did not have sufficient depth or lateral

support rotated. Concrete slabs poured on grade were

suspended in the air between the pilings.

4.5 MILTON, FL - BLACKWATER BAY

Blackwater Bay was located to the east of Pensacola

and was well east of the Hurricane Ivans eye. The

strongest winds were from the east-southeast. Wind

damage to buildings was minimal and consisted of

damage to asphalt shingles and vinyl siding. In general,

well-built homes sustained no wind damage unless

struck by falling trees. Maximum wind gusts at 10 m

(33 ft.) in open terrain were estimated to be around 45

ms-1 (100 mph). Winds were significantly lower in

forested areas. A church in town lost part of its roof

when large windows, that faced south, failed.

Storm surge ranged between 3.1 and 4 m (10 to 13ft.) above the normal water level in the bay. Buildings

on grade were gutted by floodwaters. The worst surge

damage observed occurred along Ward Basin Road and

Peterson Point Road at the north end of the bay.

4.6 NAVARRE BEACH, FL

Navarre Beach was located east of Gulf Breeze and

Pensacola Beach. The strongest winds were from the

east-southeast. Wind damage to buildings consisted of

damage to the roof coverings and vinyl siding. Some

roof sections that extended over balconies were

removed especially if they faced east or south. In

general, well-built homes sustained no wind damage.

Maximum wind gusts at 10 m (33 ft.) in open terrain

were estimated to be around 47 ms-1

(105 mph).

Storm surge ranged between 3.1 and 3.8 m (10 to 12

ft.) above the normal water level along the oceanfront

as well as along Santa Rosa Sound. Sand was

transported across the entire width of the island and

covered Rt. 399. Most of the main highway was

destroyed between Navarre Beach and Pensacola

Beach. Storm surge even crossed Rt. 98 near the bridge

leading to the island. Bridge approaches leading to the

island were eroded and had to be rebuilt before the

island was opened to local residents. Homes on grade

were severely damaged or destroyed by the stormsurge.

4.7 ORANGE BEACH, AL

Orange Beach was located just east of Gulf Shores

and experienced the east eyewall which had the highest

winds and storm surge. Buildings in town sustained

some of the greatest damage observed in our survey.

Two, five-story condominium buildings collapsed after

being undermined by the storm surge. Both were steel-

reinforced concrete structures. The storm surge ranged

between 3.8 and 4.4 m (12 and 14 ft.) above the normal

water level. Buildings were severely damaged ordestroyed especially when waves reached the second

story. Floating debris battered and mangled pilings. Up

to 2.5 m (8 ft.) of sand was removed from the beach.

Portions of Rt. 182 were washed away between Orange

Beach and Gulf Shores.

Wind damage consisted of the removal of asphalt

and metal roofing. Occasionally, roof decking was

displaced along with the roof structure especially where


6/12

they extended over a balcony. The strongest winds

appeared to be from the south with maximum three-

second wind gusts estimated to be about 54 ms-1

(120

mph) at 10m (33 ft.) in open terrain.

4.8 PACE, FL - ESCAMBIA BAY

The town of Pace was located northeast of Pensacola

at the north end of Escambia Bay. The area was east of

the eyewall and strongest winds. Wind damage to

buildings consisted of the removal of asphalt shingle

roofs and vinyl siding. Maximum wind gusts at 10m

(33ft.) in open terrain were estimated to be around 45

ms-1 (100 mph). Winds were significantly lower in

forested areas. Escambia Bay acted as funnel and

channeled the storm surge. The highest storm surge

occurred at the north end of the bay along Andrew

Jackson Dr. where still water heights reached 4 m (13

ft.) above normal water level in the bay. Waves were

superimposed on the storm surge such that the second

story floors on elevated homes even were destroyed.

Several mobile homes in Floridatown were transported

inland as much as one block.

4.9 PENSACOLA

The city of Pensacola experienced the longest

duration of strong winds from the hurricane as it was

just east of the eyewall. The strongest winds were from

the east-southeast and the highest three-second gust

reported by the Florida Coastal Monitoring Program

was just over 49 ms-1

(111 mph) at 10 m (33 ft.) in open

terrain. Winds were lower in forested areas. Wind

damage to trees was extensive and many trees fell ontobuildings. Fallen trees blocked Rt. 90, the scenic

highway. Some metal buildings sustained considerable

roof damage due to internal pressure effects when

overhead doors failed. A metal building just north of

the airport collapsed. Damage to the downtown area

was relatively minor with broken windows and

awnings. Some penthouse structures on buildings were

damaged. Light standards lost their equipment. Winds

removed some of the exterior insulation and finish

system (EIFS) siding on the Sacred Heart Hospital.

Storm surge inundated and damaged Bayfront Blvd.

and flooded a newly built subdivision just south of the

convention center. Many boats at the local marina weredestroyed. Homes on top of the bluff on the east side of

town escaped damage from the storm surge. However,

homes that were built at the base of the bluff were

gutted by the storm surge. A storm surge of up to 3.8 m

(12 ft.) even destroyed the railway at the base of the

bluff. Both lanes of I-10 were closed when portions of

the bridge deck were uplifted and removed by waves.

The bridge deck was about 4.7 m (15 ft.) above the

normal water level in the bay.

4.10 PENSACOLA - GRANDE LAGOON

One of the worst hit areas observed in our survey

was Grande Lagoon located southwest of Pensacola on

the way to Perdido Key. Most homes were constructedon grade in this low-lying area and were completely

destroyed by the storm surge. Still water heights

measured between 3.1 m and 4 m (10 and 14 ft.) with

waves superimposed on top of this level. Waves even

damaged the second story level on homes. Wind

damage to the homes consisted of the removal of

asphalt shingle roofs and vinyl siding. Maximum wind

gusts were about 53 ms-1

(120 mph) at 10m (33 ft.) in

open terrain. Winds were lower in forested areas. The

highest debris line in our survey was measured along

Seaglades Drive and was 5.6 m (18 ft.) above the water

level in Big Lagoon Sound.

4.11 PENSACOLA BEACH

Pensacola Beach was located just south of Gulf

Breeze. The Corps of Engineers had just completed a

beach nourishment project that widened the beach.

Hurricane Ivan removed much of this sand and

transported it inland burying roads and inundating

houses. Sand drifts to 3.2 m (10 ft.) were measured

around some of the homes and the sand filled several

homes. Many of the older homes built on grade were

completely destroyed by the storm surge. In certain

areas, sand was transported across the entire width of

the island. Portions of Rt. 399 east and west of

Pensacola Beach were washed away by the storm surge.The strongest winds appeared to have been from the

south-southeast. Wind damage to buildings consisted

of damaged roof coverings and vinyl siding. Wind

removed some of the exterior siding on high rise

condominiums. Some gable ends that faced south and

east blew inward. Roof structures were displaced

especially where they extended over a balcony or where

internal pressure had occurred from breakage of a

windward window. Maximum wind gusts at 10 m (33

ft) in open terrain were estimated to be around 51 ms-1

(115 mph).

4.12 PERDIDO KEY

Perdido Beach was located just east of Orange

Beach and experienced the east eyewall. The strongest

winds appeared to be from the south. Wind damage to

buildings consisted of damaged roof coverings and

vinyl siding. Wind removed some of the exterior siding

on certain high rise condominiums. Roof structures

were displaced especially where they extended over a


7/12

balcony or where internal pressure resulted from

breakage of a windward window. Maximum wind gusts

at 10 m (33 ft.) in open terrain were estimated to be

around 56 ms-1 (125 mph).

The storm surge was about 4 m (14 ft.) above the

normal water level. Buildings were severely damaged

or destroyed especially when waves reached the secondstory. A seven-story, steel-reinforced, condominium

building partially collapsed at the east end of the island

when it was undermined by storm surge. Floating debris

battered and mangled pilings on coastal homes. Up to

2.5 m (8 ft.) of sand was removed from the beach.

5. WIND VERSUS WATER

One issue in assessing hurricane damage is whether

wind or wave action or a combination of both damaged

a building. This issue arises since there are separate

insurance policies for wind and wave damage. Not

every building owner has both insurance policies.

Therefore, an accurate determination of the causes and

extent of building damage must be made. Wind and

wave forces attack a building differently. Wind forces

are greatest at roof level whereas wave forces attack the

base of the building (Fig. 9).

Figure 9. Examples of wind (a) and wave (b) damage

to housing. Relative forces are illustrated on left with

height above the ground.

Wind interacting with a building is deflected over

and around it. Positive (inward) pressures are applied

to the windward walls and try to push them down.

Therefore, it is important that a building be anchored

properly to its foundation to resist these lateral forces.

Negative (outward) pressures are applied to the side and

leeward walls. The resulting "suction" force tries to

peel away siding. Negative (uplift) pressures are

applied to the roof especially along windward eaves,

roof corners, and leeward ridges. These forces try to

uplift and remove the roof covering. The roof is

particularly susceptible to wind damage since it is the

highest building component above the ground. Wind

pressures on a building are not uniform but increase

with height above the ground and especially at roof

corners. Generally, damage to a building from windtypically begins at roof level. Thus, the last place wind

damage occurs is to the interior of the structure.

Wind damage begins with such items as television

antennas, satellite dishes, unanchored air conditioners,

wooden fences, gutters, storage sheds, carports, and

yard items. As the wind velocity increases, cladding

items on the building become susceptible to wind

damage including vinyl siding, gutters, roof coverings,

windows, and doors. Only the strongest winds can

damage the building structure. Marshall et al. (2003)

described the various failure modes in wood-framed

buildings from high winds.

Water forces are greatest at the base of the building

with a tendency to undermine foundations and destroy

support walls, thereby leading to collapse of part or all

of the building. Moving water possesses a much

greater force than that of air. A one foot tall wave

traveling at ten miles per hour possesses as much

kinetic energy as a 280 mph wind. Homes along the

coastline are at greatest risk for being damaged by

waves.

Water can lift wood buildings on pier and beam

foundations as such buildings are buoyant and can float.

Houses float landward or out to sea depending on the

ebb and flow of the water as well as the wind direction

during the hurricane. Homes with brick veneer

construction tend to rise and sink within the brickveneer shell especially if there are few or no brick ties.

Generally, these houses do not come back down to the

same position, causing distortion of the wooden-frame.

Wind does not cause this condition.

6. BUILDINGS ON CONCRETE SLABS

There were numerous buildings erected on

concrete slab foundations in the survey area. Concrete

slab foundations were either poured on-grade or

elevated by a stem wall. A stem wall involved the

construction of a concrete masonry perimeter wall built

on a concrete footing. The interior area was then filledwith dirt or sand then compacted. Most concrete slabs

measured 10 cm (4 in.) thick and contained some steel

reinforcement.

Wood-framed buildings on concrete slab

foundations were usually secured with steel anchor

bolts. These bolts were 1.3 cm (1/2 in.) in diameter and

30 cm (12 in.) long and had a J-shaped profile that

provided significant pull-out resistance in the vertical


8/12

direction. The anchor bolts were inserted into the

concrete slab when the slab was poured. Bolts had to

have sufficient height above the slab in order to pass

through the wood bottom plate and accept a steel nut

and washer. Anchor bolts were spaced 1 to 2 m (3.2 to

6.4 ft.) apart and were located within 30 cm (12 in.) of

the end of the plate and wall corners.Storm surge destroyed many buildings on slab

foundations (Fig. 10). Typically, the building frame

was removed from the slab along with most of the

contents and finish items. Occasionally, bolted wood

plates remained. The force of moving water sometimes

removed the carpeting and hardwood flooring. In some

cases, sand was scoured adjacent or beneath the slab.

The most extensive damage involved collapsing and

breaking up of the concrete slab.

Figure 10. Examples of storm surge damage to

buildings on concrete slab foundations from Hurricane

Ivan: a) cleaning, b) scouring, c) undermining, and d)

collapsing.

Buildings that were completely destroyed still left

evidence as to the direction and magnitude of the

applied forces (Fig. 11). Anchor bolts were bent along

the direction of the applied force and in some instances,

broke out of the leeward side of the concrete slab.

Nails that secured the wall studs to the wall bottom

plates also were bent along the direction of the applied

force. Copper piping was quite malleable and easily

bent. Brittle materials such as PVC and cast iron piping

frequently were broken out on the opposite side of the

applied force.

Figure 11. Indicators of direction and magnitude of the

applied force: a) bolt broken out on the leeward side of

the slab, b) broken PVC piping from an applied force

from left-to-right, c) bent nails on wood bottom plate,

and d) bent copper plumbing from a force from right-to-

left.

When destroyed buildings were encountered, other

buildings nearby were examined which survived. This

comparative analysis was done in order to determine

the height of the storm surge and resultant damage, as

well as to determine the extent of any wind damage that

might have occurred before the building was destroyed.

Wind damage to buildings on concrete slab

foundations was limited mostly to cladding items such

as roof shingles, brick masonry, vinyl siding, or

windows (Fig. 12). However, in rare instances,

portions of the roof deck were removed and gable ends

were either pushed inward or outward. Roof structures

were usually strapped to the wall top plates, and

therefore, few roofs were removed. The most

significant damage to homes from wind occurred

indirectly from trees falling on them. Trees penetrated

roofs and walls causing localized structural damage as

well as rainwater entry. Concrete slab foundations

were not damaged by wind.


9/12

Figure 12. Examples of wind damage to buildings on

concrete slab foundations after Hurricane Ivan: a)

displaced roof covering, b) uplift of roof decking, c)

removal of gable end/siding, and d) complete roof

failure.

7. BUILDINGS ON TIMBER PILES

Timber piles were either round or square and ranged

from 15 cm (6 in.) to 30 cm (12 in.) across and up to 11

m (36 ft.) deep. Piles were driven into the sand and

extended as high as 3.2 m (12 ft.) above grade. In

many instances, a concrete slab was poured on-grade

around the pilings. The slab helped stiffen the pilings

and resist soil erosion.

Wood girders were usually set into notches and

bolted to the tops of the pilings. Wood floor joists

extended perpendicular to the girders. About half the

time, floor joists were installed in the same plane as the

girders and hung by metal straps or just nailed to the

girders. In other instances, the floor joists were set on

top of the girders and were toe-nailed to the tops of the

girders or secured with metal straps. The plywood

subfloor was then nailed to the floor joists. Walls were

then erected on top of the floor. Bottom wall plates

were usually straight nailed or occasionally strapped to

the floor framing (Fig. 13).

Figure 13. Different floor details on buildings elevated

on timber piles.

Storm surge damage to buildings elevated on timber

pilings varied considerably from none to complete

destruction depending on the height of the building

above the water, depth of the pilings, and exposure to

wave action. Not surprisingly, buildings adjacent to the

coast suffered the greatest structural damage from the

storm surge. Minor damage involved removal of crossbracing between the pilings. Moderate damage

involved eroding sand around the bases of the pilings

and rotating the pilings. Concrete slabs around the

pilings were left elevated or collapsed when sand was

removed. Severe damage involved broken and crushed

pilings causing partial or complete collapse of the

building. Pilings were some times abraded or scarred

by floating debris (Fig. 14).

Figure 14. Damage to pilings from wave action: a)

breaking of cross bracing, b) erosion of sand around the

pilings, c) rotation of pilings, and d) breaking of pilings.

Floor systems exposed to wave action experienced

lateral forces from the moving water as well as uplift

forces from rolling waves. Minor damage to floor

systems involved the bowing of floor girders and/or

rotation or removal of blocking between the floor joists.

In some instances, the plywood subfloor was uplifted

causing nails to back out of the wood. Moderate

damage involved breakage of some of the floor girders

and removal of joists along the side of the building that

faced the water. Occasionally, bolts that secured the

girders were bent landward and/or upward in the

direction of the applied force. Some times, lateral wave

forces broke the floor joists or pushed them together,

stacking them toward the landward side of the building.

Uplift forces from rolling waves lifted the joists out of

their hangers. Loss of floor support led to progressive

collapse of the building with the worst damage

occurring on the side of the building that faced the

water. The effect of wave action on buildings elevated

on pilings was to dismantle them from below (Fig. 15).


10/12

Figure 15. Damage to floor systems from wave action:

a) loss of floor girders, b) rotation of blocking, c)

stacking of floor joists, and d) uplift of the plywood

subfloor.

Wind damage to buildings elevated on pilings

usually began at roof level with the loss of roofshingles, chimney caps, or antennas. In some instances,

portions of the roof deck were removed along

windward roof corners and eaves. Wind damage

progressed downward, in contrast to wave damage. In

rare instances, wind pushed the tops of frame walls

inward or outward causing the walls to rotate about

their bases. Failure occurred when straight-nailed wall

bottom plates simply pulled out of the subfloor. A

hinge formed at the base of the wall. Lack of proper

strapping and bracing contributed to such wall failures.

Floor systems were left unaffected (Fig. 16).

Figure 16. Examples of wind damage to buildings

elevated on timber pilings: a) removal of roof covering,b) gable end and roof deck failure, c) windward wall

failure, and d) nailed wall bottom plate pulled out of the

floor.

8. TORNADOES

The author encountered a number of people who

believe that buildings exploded from the low barometric

pressure in a tornado, when actually, the buildings were

gutted by the storm surge. Another myth was that

twisted trees indicated rotating winds when actually, the

trees twisted in straight-lined winds. Minor (1982),

Minor et al. (1993), and Marshall (1993) have

addressed many of these myths.

Where buildings had floated, some people believedthat houses were picked up and set back down like in

the movie Wizard of Oz. However, an examination of

these homes usually revealed pictures were still

hanging on the walls, and glassware was standing

upright in cabinets. This indicated that the houses

moved slowly (low velocity) and came to rest slowly

(low impact). Wind would have broken such items if

the house moved rapidly (high velocity) and came to

rest suddenly (high impact). Numerous homes

were completely destroyed in the surge zone but

nearby trees remained upright. Some people believed

that tornadoes simply destroyed the homes while

skipping over the trees. Generally, there was a lack of

debris up in the trees (above the surge line) and trees

had not been impaled by flying debris. Damage to

homes within the surge zone was no different than the

damage to homes similarly situated along the coastline.

The reason why buildings farther inland survived was

not because a tornado skipped over them, but that the

force of moving water (and wave action) was less than

near the coast. The author found that aerial

photographs taken by NOAA (2004) and the USGS

(2004) were invaluable in delineating wind versus wave

damage zones.

Some people believed that hundreds or thousands of

tornadoes descended upon a particular county as the

eye approached, when in fact, tornadoes did not occurin their county. The public needs to understand that

hurricane winds do almost all of the wind damage.

Tornadoes are rare, even in hurricanes. Ivan did

produce tornadoes but these were well to the north and

east of the center, and occurred mostly after the eye

made landfall.

9. SUMMARY

Hurricane Ivan struck the Alabama coast and western

Florida panhandle during the late evening on September

15, 2004 and early morning on September 16, 2004.

The author rode out the storm near Pensacola, FL thenconducted aerial and ground surveys of the area.

Detailed assessments were performed to individual

buildings during the year following the hurricane. The

purpose of these surveys was to: 1) determine the

height of the storm surge, 2) acquire wind velocity data,

3) determine the timing of each, and 4) assess the

performance of buildings exposed to wind and water

effects. Particular emphasis was placed on delineating


11/12

the damage between wind and water effects.

Storm surge from Hurricane Ivan was considerable

reaching between 3.1 to 4.4 m (10 to 14 ft.) from Gulf

Shores, AL to Navarre Beach, FL. Bays and sounds

magnified the storm surge. Homes on grade were

severely damaged or destroyed as a result of storm

surge. Even homes elevated on concrete or timberpilings sustained significant damage in the highest

surge areas. Many of these buildings were constructed

too close to the ocean. Such buildings need to have

significant set back from the water. In addition, there

was a lack of dunes in front of the buildings. It should

be recognized that dunes play a significant role in

protecting the beach as well as beachfront buildings.

The highest winds occurred from Orange Beach, AL

to Perdido Key, FL where the eyewall made landfall.

Wind speed-damage correlations suggest that the

highest wind gusts were 120 mph at roof level. Lower

wind speeds occurred inland and in wooded areas.

Three-tab asphalt shingle roofs performed poorly while

buildings covered with heavier architectural shingles

and metal performed better. In general, vinyl siding

performed poorly. A better alternative would be to

install hardboard siding. Well-built homes sustained no

structural damage due to wind. These homes were built

to meet or exceed the basic design wind speed in the

Florida Building Code of 130 mph (3-second gust, open

terrain). No tornado damage was found along the

coast. Also, no areas of category 3 winds were found in

the survey.

Buildings damaged by the storm surge and wave

action were dismantled from below. In contrast,

buildings damaged by wind had the greatest damage at

roof level. The magnitudes of the forces involveddiffered between wind and wave. In general, buildings

struck by moving water sustained more damage than

that caused by wind. A number of building deficiencies

were exploited by the storm. Such deficiencies

included inadequate pile embedment and poor

attachment of walls to floors and roofs to walls.

10. ACKNOWLEDGEMENTS

The author would like to thank the cities of

Pensacola, Orange Beach, and Pensacola Beach for

help in obtaining access to the disaster area after the

storm as well as the Escambia County SheriffDepartment. In particular, thanks to Debbie Norton

with the Emergency Operations Center on Pensacola

Beach. The National Weather Service, U.S.G.S,

FEMA, Jim Leonard, and National Ocean Service

provided much of the data within this report. Skip Jiles

with the Pensacola Aviation Center was quite helpful as

the pilot during our aerial survey. C.S. Kirkpatrick,

Polly Lawler, Kay Marshall, Scott Morrison, John

Stewart and Jim Wiethorn reviewed the paper.

11. REFERENCES

ASCE Standard 7-95, 1996: Minimum Design Loads

for Buildings and Other Structures, Amer. Soc. Of CivilEngineers, 214 pp.

FCMP, 2004: Collected data Ivan, Florida Coastal

Monitoring Program. [Available from

http://users.ce.ufl.edu/~fcmp/collected_data/storms/Iva

n/towerdata.htm]

FEMA, 1989: Is it Wind or is it Water?, Federal

Emergency Management Agency Publication, 46 pp.

FEMA, 2005: Hurricane Ivan in Alabama and Florida:

Observations, Recommendations, and Technical

Guidance, Mitigation Assessment Team Report, 294pp.

Marshall, T. P., 1993: Lessons learned from analyzing

tornado damage, The Tornado: Its Structure,

Dynamics, Prediction, and Hazards. Geophysical

Monograph, 79, American Geophysical Union, p. 495-

500.

Marshall, T. P., Bunting W., and J. Wiethorn, 2003:

Procedure for assessing wind damage to wood-framed

residences, The Symposium on the Fujita Scale and

Severe Weather Damage Assessment, 83rd Annual

Meeting, Amer. Meteor. Soc., Long Beach, CA, [CD-ROM available from http://www.ametsoc.org]

McDonald, J.R., 2005: The Enhanced Fujita Scale,

[available at http://www.wind.ttu.edu/EFscale.pdf]

Mehta, K. C., J.E. Minor, and T. A. Reinhold, 1983:

Wind Speed-Damage Correlation in Hurricane Frederic,

American Society of Civil Engineers, Journal of Struct.

Engr. Vol 109, No. 1. 37-49.

Minor, J.E., 1982: Tornado Technology in Professional

Practice, Journal of the Struct. Div., Proc. Of the Amer.

Soc. Of Civil Engineers., Vol. 108, No. ST11, p. 2411-

2420.

Minor, J.E., McDonald, J.R., Mehta, K.C., 1993: The

Tornado: An Engineering-Oriented Perspective,

Southern Region, NWS SR-147, 196 pp. [Available at

http://www.srh.noaa.gov/ssd/html/techmemo.htm]
http://users.ce.ufl.edu/~fcmp/http://www.ametsoc.org/http://www.ametsoc.org/http://users.ce.ufl.edu/~fcmp/


12/12

NOAA, 2005: Hurricane Ivan Images [Available at:

http://alt.ngs.noaa.gov/ivan]

Simpson, R.H., and H. Riehl, 1981: The Hurricane and

Its Impact, Louisiana State Univ. Press, Baton Rouge,

LA, 391 pp.

Stewart, Stacy R., 2005: Hurricane Ivan Tropical

Cyclone Report., National Hurricane Center.

[Available online at:

http://www.nhc.noaa.gov/2004ivan.html?]

USGS, 2004: Coastal impact studies after Hurricane

Ivan. [Available at:

http://coastal.er.usgs.gov/hurricanes/ivan/photos]
http://alt.ngs.noaa.gov/ivanhttp://www.nhc.noaa.gov/2004ivan.htmlhttp://www.nhc.noaa.gov/2004ivan.htmlhttp://alt.ngs.noaa.gov/ivan

hurricane ivan damage survey

Documents